THEORETI CAL AND EXPERIME NTAL STUDIES OF THE REMOTE FIELD EDDY

CURRENT EFFECT

S. Nath. W. Lord andY. S. Sun*

Departmen t of Electrical Engineering
Colorado State University
Fort Collins. CO 80523

INTRODUC TION
Convention al eddy current (EC) nondestruct ive testing methods. based on the
principle of electromagn etic induction have been used over the years for defect detection. A
low frequency alternating current is fed to the excitation coil which induces eddy currents
in the material. These eddy currents generate a magnetic field. inducing a secondary voltage
in the sensor coil. The change in the impedance of the sensor coil is used to indicate the
presence of material inhomogene ities (Figure 1 ). Figure 2 shows a conventiona l eddy
current probe and a remote field eddy current (RFEC) probe. The difference between the
t w o is that in the convention al EC probe the excit er and the sensor are t he same coil.
normally operating between 1KHz and 10 MHz. while in the RFEC probe the exciter and
the sensor coils are several pipe diameters apart with a frequency of operation from 40 to
160Hz. The RFEC probe does not measure the change in the impedance of the sensor coil,
rather the steady state A.C phase angle difference between the exciter and the sensor is
monitored (Figure 3). Skin effect characterist ics limit the eddy currents to the surface of
the material. restricting the conventiona l EC testing methods to the detection of near
surface inhomogeneities. The RFEC probe in contrast overcomes the above limitation due to
its apparent sensitivity to both inner and outer diameter (ID and OD) defects [1]. Recently
numerical models have been developed to develop a better understand ing of the physics of
the phenomeno n [2-8].

z1 z 2

T;:s; OBJEC7

Fig 1. Convention al eddy current NDT test setup measuring the probe impedance.

267
 Pipe Wall Th ickn ess

Exciter Coil Sensor Coil

Fig2. Conventional EC and RFEC probe.

~ ~ FRO"
SENSOR ---i AAIPUf1ER
D<C:TER SENSOR COIL
CCIL COIL

I
I
~
SICNA!. SOURCE: ST""<.l'P::R lolOTOR PHASE: LOCK-IN
HP 816.5A
OR!Vc: SYST!'~ AJJI=>UfiE~
&< ltnoco J 961 1
:00 w
I
>l.IPUi1E~

I r
I
ISM s rs.- EM gooo
I
fREQUENCY ANO c;.;N £2..E~T
STI;?PE:R "OTOR CONTROL
OATA LCGCINC

Fig 3. Experimental test rig block diagram to monitor the steady state AC phase angle
difference between the exciter and sensor coils of a RFEC probe.

268
FINITE ELEMENT MODEL
The quasi-static form of Maxwell's equations describe the electromagnetic induction
phenomenon characteristics of both the conventional EC and the RFEC effect. Because of
the negligible displacement current term, the quasi-static equations reduce to

(1)

where A is the magnetic vector potential. J. is the source current density. w is the angular
frequency. cr is the electrical conductivity, fJ.. is the magnetic permeability.
Finite element (FE) techniques have been successfully applied to model E.C
phenomena for different geometries [9-11]. An axisymmetric finite element model has been
developed to study the basic RFEC phenomenon [3]. This model has been further extended
to study the effect of parameter variations on the RFEC phenomenon. Information on the
details of the model and the basic physics of the phenomenon is available in references [2-
4]. Space limitations preclude additional information in this paper.

FINITE ELEMENT PREDICTIONS

Figure 4a) compares the experimental finite element prediction for the RFEC probe
magnitude and phase without a defect and Figure 4b) shows the phase characteristics for an
axisymmetric defect at 40Hz. The results confirm the validity of the FE approach .

.•
1'------~·"·.!!...."- - - - r ..
LO

-- • "'~
- -.• c;5'
-:-t..J :I.
l

g
- 1.5
.,
0
-Z . .J
2
1:;
-2-~ "Q
C)
•l:. Q g,
-:J. , ,..,
1 ..,
i .. .. ....
.
3:::

Fig 4a. Experimental and FE predictions for the RFEC probe magnitude and phase
without a defect at 40Hz.

l
,. -··- nNt•
-
.:.ment
·--".rl"'"""'tol

I!;
...
:§ tSO \

i
;;: I
\
\ '\
I

Ef --- ,
'
' - -----

110

too
-t. s '
t.o ·
-.1 ...
'

Fig 4b. Experimental and FE prediction for the RFEC probe phase characteristics with
defect at 40Hz.

269
 The parameter variation studies include frequency characteristics. pipe wall thickness.
pipe inner diameter. and coil width and depth variations. There is also a comparison of the
flux plots from an ID and OD defect.
Figures 5 to 10 illustrates the phase and magnitude prediction for the RFEC probe. It
is obvious that the probe is sensitive to higher frequencies. magnitude attenuation is higher
with increasing pipe wall thickness and pipe inner diameter. The exciter coil dimension
variations indicate a negligible change in the RFEC characteristics. Figures 11 and 12 show
the instantaneous flux contours and phase characteristics for an ID and an OD defect. The
plots confirm that the probe is sensitive to both ID and OD defects.

COIL WIOTH OEPENDENCE: Mo gn l l ude

12. 0

11. 5
t. Call Wt . • O.ZS"
11.0 2. eou we . • o.SO"'
), Co il Vt . • 1.00"'
4. Co il \H , ., 2.00"
10. 5 ), CoU Wt. • 4 . 00"'
!!; 6. eou wc . .. s.oo"'

a
ID • l.O"
10.0

!h
9. 5

9. 0

i
'6
e. 5

8. 0
"'
:§' 7 .5

7. 0

6 .5

6. 0
0 2 3 s 8 7 8 6 10
OJ~t.onee Do wn Pl p e (C Jol'llet..er"l

Fig 5 FE prediction of probe magnitude characteristics for exciter coil width variations.

COI L DIAMETER OEPENOENCE : Magnllu de

IZ r-~-r--~,--r-,--~-r-,--r-~-.--r-,r~--r-~-r--.-,

11
1. (I - ].0"
2. od • 2.7S"
). d - 2.)"

l5 10
t. .
~.
4 ,. 2. 2)"
d - 2.00..
o. d•l.J!t"'

O J :~~~t.onc:e Oown P Jpc:[ O i o i'U!t.er .s l

Fig 6 FE prediction of probe magnitude characteristics for excit er coil diameter

variations.

270
 FREQUENCY DEPENOENCE:Phose Angle
200
10 H:r..

ISO 160Hz.

l!i 100
20 Hz.

8
!b
50
40 H:r. .

I 0

..
'a
-so
:t 80 R2; .

1-100

-150

-200
0 2 3 5 6 7 8 10
018\..ance Oown Plpe (01a!ftet.er&)

Fig 7 FE prediction of probe phase characteristics for various excitation frequencies.

Frequenc~ Ct\ar-act.erl3l1cs wllh derect.

Fig 8. FE prediction of phase characteristics for various excitation f requencies with a

defect at 40Hz.
INNER DIAMETER DEPENCENCE:Mognllude
14 --r--r-~

IS d . 48"

....
1.
L I!. 36"
) . 4 • . 24"
d. 16"
12
'·
....'·
4 .. 1 2"
4 .. 6"
d .. 3"
d .. 1. 5"
4 .. 0.5"

l-5 6 1 8

01at.ance Down J-lpaiDJa"'a\.ersl

Fig 9 FE prediction of probe magnitude characteristics for pipe inner diameter

variations.

27,
 WA ~~ -THJCKNESS O EPENOENCE :M ognl~ude

12

11

J 1 I ! t I

0 1 3 i 'S c 7 10
OJ~t.onc e Down PJ.~eiOJOI'II.I!!Iler-a)

Fig 10 FE prediction of probe magnitude characteristics for various pipe wall thickness.

Fig lla. FE prediction of instantaneous fiux contours for a ID defect.

Fig llb. FE prediction of instantaneous fiux contours for a OD def ect.

272
 I . D end 0 . 0 DEFECT - PHR SE

• · ~- J . 0 . Dt f1c I
-110 o . o. o.t..:t

- 120
:s
8 -130

!
!. -1~0

b
]i
'0 - 150

~-HiD
I
- 170 ~

Fig 12. FE prediction of phase characteristics for aID and OD defect at 40Hz.

Differential eddy current probes have been used extensively to test steam generator
tubing in a number of industries [11.12] and it has been shown that the corresponding
impedance plane trajectories can characterize defects. A similar approach has been used to
plot the RFEC probe characteristics [ 13]. A single sensor coil conventional RFEC probe was
used to simulate a differential signal. The differential signal y(n) can be synthesized from
the absolute RFEC signal x(n) using the relation y(n) = x(n)-x(n-N). The distance between
the differential sensor coils d is given by d = Nd'. where N is an integer number and d' is the
incremental distance moved by the probe. If the real and imaginary component of the
sensor coil induced voltage are plotted in the complex plane. a closed trajectory is obtained.
analogous to the conventional EC complex impedance plane trajectory (Figure 13).

CONCLUSIONS
The RFEC probe characteristics is governed by the quasi-static form of the Maxwell's
equations and so can be modeled by conventional eddy current finite element code thus
confirming that the RFEC phenomenon is a diffusion process strictly governed by the
parabolic diffusion equation. The parameter variations studied show the capability of the
finite element code to help in understanding the physics of the phenomenon. and also serve
as a useful tool for probe design. This study also confirms the RFEC probe sensitivity to
both inner and outer diameter pipe wall defects. Defect characterization using the RFEC
trajectories scheme is a novel and useful technique. Furthermore. the finite element code
can be easily extended to larger pipe diameters.

varJcOlo Cefect Wl dlh - T~ oJ ~clorlo~

,,,
]
I I

:f j
;'t
;:.

~
'·

;:
Q

]
-2 . :1 '
•2 .0 '
-t.S .. l .O ...
Vcosa. - v c o ••.l
..' '
t.O ... z.o
j...
Fig 13 RFEC trajectories.

273
ACKNOWLEDGMENTS
This research has been supported by the American Gas Association. The authors
would like to thank all the researchers in the NDE laboratory at Colorado State University
for the numerous discussions and help.

REFERENCES
T. R. Schmidt. "The remote field eddy current inspection techniques," Material
Evaluation, Vol. 42, Feb 1984. pp. 225-230.
2 W. Lord. et al., "A finite element study of the remote field eddy current
phenomenon:· IEEE Transactions on Magnetics. Vol. 24, Jan 1988, pp. 435-438.
3 W. Lord. ··Final report on the finite element model of the remote field eddy current
effect:· Project PR 179-520, Sept 1986, American Gas Association.
4 W. Lord. et al.. "Physics of the remote field eddy current effect:· Review of Progress
in Quantitative NDE , D. 0. Thompson and D. E. Chimenti Eds., Plenum Press, New
York 1988, pp. 165-172.
5 H. Hoshikawa. et al., ··study of the remote field eddy current testing:·
Electromagnetomecha nicallnteractions in Deformable Solids and Structures . edited
by Y. Yamamoto and K. Miya. North Holland Press. Amsterdam 1986. pp. 215-220.
6 T. R. Schmidt. et al., ··Experience with the remote field eddy current technique:·
Proceedings of the Third National Seminar on Nondestructive Evaluation of
Ferromagnetic Materials", March 23-24. 1988, Houston, Texas. pp. 85-107.
7 D. L. Atherton. et al.. "Finite element calculations of the remote field through wall
eddy current inspection tool for nuclear pressure tubes:· Proceedings Vol. 1.
Workshop on Electromagnetic Field Computation . Schenectady, Dec 1986.
8 D. L. Atherton. S. Sullivan. ·The remote field through wall electromagnetic
inspection technique for pressure tubes:· Materials Evaluation, Vol. 44. Dec 1986,
pp. 1544-1550.
9 W. Lord. ··Applications of numerical field modeling to electromagnetic methods of
nondestructive testing:· IEEE Transactions on Magnetics, Vol. 19, 1983, pp. 2437-
2442.
10 R. Palanisamy and W. Lord. ··Finite element analysis of eddy current phenomenon:·
Materials Evaluation, Vol. 18 , 1980, pp. 39-41.
11 N. Ida and W. Lord, "Graphical simulation of electromagnetic NDT probe fields:·
IEEE Computer Graphics Applications. 1983, pp. 21-28.
12 R. Palanisamy and W. Lord. ·· Finite element simulation of support plate and tube
defect eddy current signals in steam generator NDT:· Materials Evaluation,
Vol. 39, June 1981, pp. 651-655 .
13 S. S. Udpa. et al.. ··signal processing for remote field eddy current inspection
technique:· Review of Progress in Quantitative NDE, D. 0. Thompson and D. E.
Chimenti Eds. Plenum Press. New York, 1988.

• Professor Y. S. Sun is a visiting faculty from the Nanjing Areronautical Institute, PRC.

274